Search

A supercomputer may provide a heterogeneous environment including the host processors and accelerators. As a consequence, programmers can execute applications on both host processors and accelerators. To achieve the combined performance potential, it requires software to effectively partition the the workload of parallel applications to maximize the computation overlap between host processors and accelerators. However, it is hard to determine the right data partition and task parallelism on heterogeneous platforms given a new application. The number of possible options regarding data partition between host processors and accelerators is huge. The imbalanced data and task partition can seriously hurt the performance. In this paper, we present an approach to determining the workload partition and the task granularity for any given application, targeting the Intel Xeon Phi accelerated heterogeneous systems. We employ machine learning techniques to train a predictive model off-line and then use the trained model to predict the data partition and task granularity for any unseen programs at runtime. We apply our approach to 21 representative parallel applications and evaluate it on a Xeon-Xeon Phi mixed heterogeneous platform. Compared with the optimized, non-asynchronous method, our approach achieves, on average, a speedup of 1.6× and 1.9× using a single MIC with one CPU processor and two CPU processors.

In this paper, we propose an end to end model for keyword spotting(KWS) with densely connected convolutional network(DenseNet) and the integration of Short Time Fourier Transform(STFT) which aims to extract utterances from raw waveform directly. Furthermore, we investigate the efficiency of adaptive and trainable window function in the task of keywords spotting. Using the recently-released Google Speech Commands Dataset as our benchmark. Our DenseNet implementation significantly outperforms previous neural networks based KWS system in terms of accuracy.

Coprocessors based on Intel Many Integrated Core (MIC) Architecture have been adopted in many high-performance computer clusters. Typical parallel programming models, such as MPI and OpenMP, are supported on MIC processors to achieve the parallelism. In this work, we conduct a detailed study on the performance and scalability of the MIC processors under different programming models using the Beacon computer cluster. Followings are our findings. (1) The native MPI programming model on the MIC processors is typically better than the offload programming model, which offloads the workload to MIC cores using OpenMP, on Beacon computer cluster. (2) On top of the native MPI programming model, multithreading inside each MPI process can further improve the performance for parallel applications on computer clusters with MIC coprocessors. (3) Given a fixed number of MPI processes, it is a good strategy to schedule these MPI processes to as few MIC processors as possible to reduce the cross-processor communication overhead. (4) The hybrid MPI programming model, in which data processing is distributed to both MIC cores and CPU cores, can outperform the native MPI programming model.

Accelerators have become critical in the process to develop supercomputers with exascale computing capability. In this work, we examine the potential of two latest acceleration technologies, Nvidia K20 Kepler GPU and Intel Many Integrated Core (MIC) Architecture, for accelerating geospatial applications. We first apply a set of benchmarks under 3 different configurations, i.e., MPI+CPU, MPI+GPU, and MPI+MIC. This set of benchmarks include embarrassingly parallel application, loosely communicating application, and intensely communicating application. It is found that the straightforward MPI implementation on MIC cores can achieve the same amount of performance speedup as hybrid MPI+GPU implementation when the same number of processors are used. Further, we demonstrate the potentials of hardware accelerators for advancing the scientific research using an urban sprawl simulation application. The parallel implementation of the urban sprawl simulation using 16 Tesla M2090 GPUs can realize a 155x speedup compared with the single-node implementation, while achieving a good strong scalability.

In light of the six lexical cohesive ties that do not necessarily satisfy the additive situation, a new method is proposed to evaluate lexical cohesion based on Sugeno measure. The evaluation indexes are classified anew and weight parameters are given by experts. Then Sugeno measure is applied to integrate them. The experimental results show that the integrated results are significantly correlated with teachers' scoring of the compositions, which indicates the effective and comprehensive quality evaluation of lexical cohesion in English writing.

Spam is unsolicited bulk email. Most of the popular spam filters work depending on individual training corpus or discrete feature extraction, and these filters haven't considered spam's bulk feature. The better method to stop spam may be the collaborative approach that does not only consider the content of the email but depends on the collaboration of groups of users who share information about spam. Therefore, this paper proposes a spam filtering technique based on feature fingerprint in multi-recipient collaboration.

A contention-based geographic routing (CGR) protocol for mobile ad hoc networks is proposed, which operates in two schemes: greedy-forwarding scheme and void-handling scheme. The greedy-forwarding scheme consists of basic mode and optional mode. In the basic mode, when receiving a data packet from upstream node, intermediate nodes within its forwarding area contend to relay it by waiting for a dynamic forwarding delay (DFD), which is computed based on its own position relative to the upstream node and destination node. The intermediate node with minimum DFD transmits the data packet first and other nodes cancel their scheduled transmission for the same data packet by overhearing its packet forwarding. In the optional mode, upstream node can send data packets to its specified downstream node as long as the communication link between them is valid. An accurate link forecast scheme is used to compute the valid time based on location and mobility information of upstream and downstream nodes. In the void- handling scheme, a one-hop flooding method is introduced to overcome communication void problem. Finally, simulation results show that CGR outperforms LAR1 and BLR protocols in terms of end-to-end delay and packet delivery ratio.

A contention-based beaconless geographic routing (CBGR) protocol for mobile ad hoc networks is proposed, which includes a basic mode and an optional mode. In the basic mode, when receiving a data packet from upstream node, intermediate nodes in the forwarding area contend to relay it by waiting a dynamic forwarding delay (DFD), which is computed based on its relative position with the upstream node and destination node. The intermediate node with minimum DFD transmits the data packet first and other nodes cancel their scheduled transmission for the same data packet by overhearing its packet forwarding. In the optional mode, upstream node can send data packets to its specified downstream node as long as the communication link between them is valid. An accurate link forecast scheme is used to calculate the valid time based on location and mobility information of upstream node and downstream node. In addition, a one-hop flooding method is introduced to overcome communication void problem. Finally, simulation results show that the CBGR protocol outperforms the LAR1 and BLR protocols in terms of end-to-end delay and packet delivery ratio.